The present invention is related to a client-server based information providing system. In particular, the present invention provides a method of searching for information on the basis of a level of similarity of the information to a primary document.
Finding documents of interest on a server whether connected to the Internet or not has become extremely challenging because of the abundance and diversity of the available documents. Traditionally, two approaches have been available to users seeking documents of interest: index-like search engines and search engines.
Indexes transmit information to a client computer in the form of a Web page comprising document categories such as arts, sports, finance, etc. When a user selects a category, such as by clicking on a button corresponding to the category, one or more subcategories are transmitted to the client computer in a second Web page. Thus, for example, if a user selects xe2x80x9cfinancexe2x80x9d, subcategories comprising equities, interest rates, and corporate summaries might be transmitted with the second Web page.
Index-like search engines, however, often fail to provide a user with a document of interest. For example, indexes require the construction of a structured database comprising a plurality of documents that are categorized with respect to the information therein. The need for operators to enter the documents into the database disadvantageously limits the size of the database and accordingly, the number of documents that can be found using the index-like search engine. Therefore, a document sought by a user may not be contained in the index-like search engine.
Search engines, on the other hand, typically do not utilize structured databases to provide information to users having a query. Instead, a user must provide a search expression comprising one or more search terms, such as a key word that the user perceives as relevant to his/her query. Subsequently, a database comprising a plurality of documents is searched for matching documents that partially or completely match the search expression received from the user. Links to one or more of the matching documents is then transmitted to a client computer in the form of a Web page hypertext buttons. A user may select or click on one of the links to either go to the selected document or obtain additional information from the server about the document.
For a number of reasons, however, traditional search engines often fail to retrieve a document that comprises information that satisfies a user""s query. For example, because traditional search engines require the receipt of at least one specific search term from the user, a user who is unfamiliar with such a search engine may be unable to provide a response to the search engine""s request for a search expression either because they do not know how or become intimidated. Even if a search expression is received from such a user, the search expression may comprise nonspecific or extraneous terms that prevent the search engine from retrieving a selection of pertinent documents. Experienced users also may have difficulty utilizing a traditional search engine. For example, users often have only an intuitive idea of the search terms that actually characterize the sought after document and are, therefore, unable to provide the specific search terms needed to pose a successful query to the search engine.
What is needed, therefore, is a method of providing information to a user who is unable to supply specific search terms characterizing the information he/she seeks.
The present invention relates to a method for searching and providing information to a user. The method includes the steps of transmitting a primary selection to the user, receiving a level of interest in the primary selection from the user and obtaining a set of associated selections having a level of similarity to the primary selection determined on the basis of the level of interest, and transmitting at least a subset of the associated selections to the user.
Preferably, the obtaining step comprises searching a database to obtain the set of associated information selections. The level of similarity is preferably determined to have one of at least three different levels and a different set of associated selections is obtained depending on which level is determined.
Typically, the primary selection and at least some members, for example, each member, of the set of associated selections are characterized by a plurality of attributes and the level of similarity is determined from the absolute or relative number of attributes common to a member of the set of associated selections and the primary selection. The attributes preferably comprise at least one of text related to the associated selections or primary selection and a graphic related to the associated selections or primary selection. For example, the attributes may comprise at least one of a word, a phrase, and a keyword.
In a preferred embodiment, the primary selection relates to a primary item and each member of the set of associated selections relates to a different associated item. For example, the primary item and the associated items preferably relate to at least one of a document comprising information, an object for sale, an object for rent, and a service. The document preferably comprises at least one of HTML (hypertext markup language, DHTML (dynamic hypertext markup language), x-text, or x-HTML (extensible markup language). The object for sale preferably comprises at least one of computer software, a compact disc, a video, a book, an article of food, an automobile, a source of audio provided over the Internet, a source of video provided over the Internet, and an article of clothing. The object for rent preferably comprises at least one of a compact disc, a video, a book, or an automobile.
In a preferred embodiment, the method further comprises receiving a new primary selection from the user and replacing the primary selection with the new primary selection. The new primary selection is preferably a subset of the set of associated selections. The method may further comprise optionally receiving a new level of interest from the user and replacing the level of interest with the new level of interest, searching a database to obtain a new set of associated selections having a level of similarity to the new primary selection determined on the basis of the level of interest and transmitting the new set of associated selections to the user. The optional receiving, searching, and transmitting the subset steps are repeated in a preferred embodiment.
The present invention also relates to a computer system for delivering information to a client computer, comprising a server configured to transmit a primary selection to the user, receive a level of interest in the primary selection from the user, obtain a set of associated selections having a level of similarity to the primary determined on the basis of the level of interest, and transmit at least some of the associated selections to the user.
The system is preferably configured to search a database to obtain the set of associated information selections. In a preferred embodiment, the server computer is configured to partition the level of similarity into one of at least three different levels and obtain a different set of associated selections depending on which level is determined. The primary selection and each member of the set of associated selections are preferably characterized by a plurality of attributes and the level of similarity is determined from the absolute or relative number of attributes common to a member of the set of associated selections and the primary selection. The attributes preferably comprise at least one of text related to the associated selections or primary selection and a graphic related to the associated selections or primary selection. For example, the attributes may comprise at least one of a word, a phrase, and a keyword.
In a preferred embodiment, the primary selection relates to a primary item and each member of the set of associated selections relates to a different associated item. For example, the primary item and the associated items relate to at least one of a document comprising information, an object for sale, an object for rent, and a service. The document preferably comprises at least one of HTML (hypertext markup language, DHTML (dynamic hypertext markup language), x-text, or x-HTML (Extensible Markup Language). The object for sale preferably comprises at least one of computer software, a compact disc, a video, a book, an article of food, an automobile, a source of audio provided over the Internet, a source of video provided over the Internet, and an article of clothing. The object for rent comprises at least one of a compact disc, a video, a book, or an automobile.
In another preferred embodiment, the system further comprises software to receive a new primary selection from the user and replace the primary selection with the new primary selection. Preferably, the new primary selection is a subset of the set of associated selections. The server is preferably configured to, optionally receive a new level of interest from the user and replace the level of interest with the new level of interest, search a database to obtain a new set of associated selections having a level of similarity to the new primary selection determined on the basis of the level of interest, and transmit the new set of associated information selections to the user.